NOvITATES

AMER]ICAN MUSEUM
NO vITA TES
PUBLISHED BY THE AMERICAN MUSEUM OF NATURAL HISTORY
NUMBER 1426
SEPTEMBER 13, 1949
CITY OF NEW YORK
IDENTIFICATION AND SIGNIFICANCE OF
THE CUCURBIT MATERIALS FROM
HUACA PRIETA, PERU
BY THOMAS W. WHITAKER1 AND JUNIUS B. BIRD
CONTENTS
THu ARCHAEOLOGICAL BACKGROUND, BY JUNIUS B. BIRD ...............
1
Utilization of Gourds ............................................
Containers ....................................................
Fish-net Floats .....................
...........................
Discs ........................................................
..............................................
Scoops or Ladles .
............................
Decorative Techniques .............
Repair of Broken Gourds .......................................
Material Associated With Early Ceramics ..........................
Summary .......................................................
BOTANICAL IDENTIFICATION AND SIGNIFICANCE, BY THOMAS W. WHITAKER.
Methods .......................................................
Materials .......................................................
Discussion and Conclusions .......................................
3
3
5
5
5
5
6
6
7
8
8
9
12
LITERATURE CITED ................................................
15
THE ARCHAEOLOGICAL BACKGROUND2
The cucurbit materials described by Dr. Whitaker were recovered at a site known locally as the Huaca Prieta on the coast line
of the Chicama Valley, Peru, at about latitude 8° S. Despite
severe erosion, the huaca, a large compact midden, still has a
1 Geneticist, Bureau of Plant Industry, Soils, and Agricultural Engineering, United
States Department of Agriculture, La Jolla, California. This work was completed
while the author was a Fellow of the John Simon Guggenheim Memorial Foundation.
2 This field work was a part of the Andean Institute-Virni Valley survey, a cooperative undertaking shared by the American Museum of Natural History, the Chicago
Natural History Museum, the Smithsonian Institution, and Columbia and Yale
Universities, 1946-1947.
2
AMERICAN MUSEUM NOVITATES
NO. 1426
maximum thickness of 45 feet. To determine its nature and
contents, blocks or columns of debris were isolated and then
separated into successive layers by following the visibly traceable strata. In one pit (HP-3) 22 such layers, some subdivided
into several parts, spanned the full period of occupation of the
site. As study of the collection proceeds, it appears from the
distribution of certain items that the material of these layers can
be grouped into six or seven chronological units.
A second pit (HP-2) provided additional material roughly
contemporaneous with the lower six levels or first two units of
the larger section. In both excavations the dirt was screened,
and all plant remains and artifacts were saved. Because of ideal
conditions for preservation, the quantity of perishable material
exceeded all expectations. As an example, the squash and gourd
remains from these two pits amounted to over 11,0001 shell fragments, 1300 seeds, and about 550 peduncle fragments. Of this
pre-ceramic material only that from Layer M in HP-3 was sent to
Dr. Whitaker for study. This did not include 1500 shell fragments from this layer discarded after the field record was made.
The series for study was selected for various reasons: its position
well down in the mound near the lower limits of best preservation,
the high yield of this particular layer, and a hope that it would
prove to be an adequate sample, since it apparently contained examples of everything found in the other layers.
As preliminary reports (Bennett, 1948; Bird, 1948) on this
excavation have appeared and as a monograph is in preparation,
my comments are li-mited to details that explain the setting and
supplement Dr. Whitaker's study.
From the excavations it was clear that the concentration of
debris was the result of continuous occupation of a limited area by
a people whose economy was based partially on agriculture, partially on fishing. Their material culture was in some respects
quite primitive. In view of the subsequent ceramic record in
Peru, it is of general interest that they had no knowledge of
pottery. This and other evidence indicate considerable antiquity
for the remains, prompting a guess that we are dealing with a culture period which may fall between 3000 and 1000 B.C.
The plants utilized during this period, except for the Cucur1 In order to compensate for the breakage resulting from excavation and handling,
the small shell fragments showing fresh fractures were grouped by three's and counted
as the equivalent of one unbroken piece.
1949
CUCURBIT MATERIALS FROM PERU
3
bitaceae, have been tentatively identified. Beans of at least
three varieties (all Canavalia?), chili peppers, cotton, and canna
were probably cultivated. Evidence for the latter consists principally of many scraps of the rhizome skin and some leaves.
Definitely wild, but commonly utilized as food, were the small
tubers from Scirpus (americanus?) and Cyperus sp.? Roots of
cat-tail seem also to have been eaten. Important fruits were the
Lucuma and Bunchosia sp.? Two other plants, as yet unidentified, were of considerable economic importance: one provided
bark for barkcloth, and the second, possibly Agave, yielded a bast
fiber used to some extent in combination with cotton in the production of many fabrics. In this period there was no evidence
for the presence of maize.
UTILIZATION OF GOURDS
Among the 10,770 gourd shell pieces for which we have relative
position data in HP-3, only 358 (3.3 per cent) give some clue as to
their use. An additional hundred bear some slight indication of
having been cut or worked in some manner. This percentage
is so low that it might well be misleading to limit the discussion
to the gourd shells found with the Layer M material. Thus
the following is based on an analysis of the 358 specimens which
are classifiable into four categories: containers, 72.4 per cent;
fish-net floats and float stoppers, 15.1 per cent; discs of unknown
use, 5 per cent; and fragments of scoops or ladles, 7.5 per cent.
CONTAINERS: Among these, open-mouthed bowls predominate,
all made by sectioning the gourd transversely. This, unlike a
longitudinal cut, yields only one container per gourd. Usually
the incision was made high enough so the bowl rims were incurving; very few were cut where the fruit reaches its maximum diameter. One rim fragment of this latter type, found in the top
level of the pre-ceramic debris, has an indicated diameter of 40
to 42 cm. Another, found midway between this and the M layer,
apparently measured 34 cm. in diameter. These sizes are exceptional; as a rule the bowls were under 22 cm. in diameter, with a
possible depth of 10 or 12 cm.
The short-necked type of Lagenaria siceraria, with the neck
cut off, produced a constricted-mouth jar form. These vary considerably in size and rim profile. Cuts made farther utp on the
neck of the long-necked L. siceraria yielded a bottle form, but it
is impossible to distinguish with certainty whether the fragments
4
NO. 1426
AMERICAN MUSEUM NOVITATES
are from these bottle forms or from fish-net floats. The same
type of gourd served as small spherical containers about 6 cm. in
diameter. Remains of two of these had been fitted with flattened
circular covers cut from a much larger specimen.
.
.
FIG. 1. Fruits of Lagenatriat siceraria from Huaca Prieta. A. Virtually intact fruit, Test 5; probably used for fish-net float. B. Fruit of L. siceraria,
Test 3; used for fish-net float.
The only other objects here classed as containers are a few
fragments, also of the long-necked Lasgencaria sicercarica, which have
had part of one side removed to form a dipper. The relative
frequencies of these and the other types of containers are as
follows:
1949
5
CUCURBIT MATERIALS FROM PERU
Bowls, incurving rims
Bowls, flaring or vertical rims
Jar form, constricted mouth
Bottle form
Small spherical containers
Dippers
PER CENT
80.4
6.9
5.4
4.3
1.1
1.9
FISH-NET FLOATS: The fortunate discovery of a large fish net
with eight floats attached and still intact positively identifies
these objects. All have the tip of the neck cut off and the seeds
removed, though a few still remained in the one sent to Dr.
Whitaker (figs. iB, 3C). The openings are plugged with gourd
shell discs from 10 to 22 mm. in diameter. Such small discs
occur from top to bottom of the mound, sometimes still in place
in a float neck fragment. Gourd floats are still used along this
coast, but none of the long-necked forms were observed nor are
the seeds removed as in the past.
DISCS: Discs larger than those used in the floats occur in the
upper third of the midden. Plain, smooth-edged discs vary in
diameter from 3 to 7 cm. Rough-edged discs with one, two, or
three perforations near the margin are slightly larger. Another
well-finished example has 12 small punctures at the margin and
four holes at the center in which cords remain.
Scoops OR LADLES: Various fragments of roughly oval and
angular objects with worn or smoothed edges can perhaps be
classed as ladles. The absence of complete specimens and the
lack of uniformity among the fragments make a clear description
difficult. Apparently any broken gourd fragment of sufficient
size was suitable as a makeshift spoon for stirring or serving food.
One occasionally sees such pieces in use today and though they are
not cut to a specific pattern they are nonetheless real utensils.
DECORATIVE TECHNIQUES: In the entire lot of shell fragments,
all from containers, only 13, or one in 830, were decorated. With
such a low rate of occurrence, it is surprising to find several distinct methods of decoration.
Crudest are those with lines and marks made by scratching off
the epidermis when it was fresh and soft. In nearly all the
incurved rim bowls, the epidermis has been removed just below
the lip for a distance varying from 3 to 15 mm., but I hesitate to
class this as a decorative feature. Fine line incising in the hard
shell was used for diagonal crosshatching in two instances. Five
6
AMERICAN MUSEUM NOVITATES
NO. 1426
pieces show pyro-engraving in tantalizingly incomplete fragments
of decoration. Two show rather skilful carving of geometric
stylized faces. In these, both the design and the quality of work
lead us to believe that they are not casual or experimental products, but rather that the method of decoration may have some
traditional background. If this is true, then the rarity of decoration is still more of a puzzle.
REPAIR OF BROKEN GOURDS: Broken containers were sometimes repaired by lacing the cracks with cotton cord. Examples
are so poorly executed as to suggest that gourds were so abundant
that skilful repairing was unnecessary.
MATERIAL ASSOCIATED WITH EARLY CERAMICS
After the work in the main mound was completed, additional
excavations in a smaller accumulation immediately to the north
exposed more pre-ceramic debris underlying and blending with
pottery-bearing refuse. This provided sufficient material, both
botanical and cultural, to establish the relation of the pre-ceramic
horizon with the known Cupisnique-Chavin culture. The latter
has been described on the basis of the more durable artifacts, and
little else relating to it has been found. The nearly simultaneous
appearance for the first time in this part of Peru of several plants,
maize, peanuts, avocados, inga, and a "warty" squash, together
with new textile techniques and the Cupisnique ceramics, implies
the arrival here of migrants from another region. Items marked
HP-5 are from this period and were removed from a considerably
smaller volume of debris than that examined in HP-3. Although
over 16,500 fragments of pottery were taken from it, the yield
of squash and gourd remains was small: 1721 shell fragments, of
which 137 are from the warty squash, 144 peduncles, and perhaps
a hundred seeds.
The analysis of the gourd shell pieces compared with the preceramic material reveals some distinctions. Diagnostic specimens, that is, those of determinable use, constitute 5.8 per cent of
the total. The proportion of decorated fragments is much higher,
one in every 28, or 3.5 per cent of the total. The only method
represented is pyro-engraving, with which a variety of curvilinear
designs was created. In addition to the bowl forms already mentioned, there are three fragments from bowls made by cutting
the fruit longitudinally. The bottle form containers are represented, but the small spherical type, the dippers, and ladles were
1949
CUCURBIT MATERIALS FROM PERU
7
not found in this lot. Unique is a whistle made from a nearly
spherical shell, 4 mm. thick, 45 mm. long, and 47 mm. in diameter
(fig. 2). The seeds had been removed by cutting out the peduncle
11 11 ^<- c," °>-~ ~ ..c~ .-~:N;. . .
ll | , E, S, ,j., p, , t, .M,z,., E, ,i, , , ,n.,n.,. . .. . . .. ~. ~.~.~ ~ ~ ~ .
y..
:.:..
...
FIG. 2. Spherical fruit of Lagenaria siceraria made into whistle. Found
with Cupisnique ceramics and early maize (HP-5-A3).
scar and after the insertion of a round clay (?) pellet the hole
was closed with a gourd disc. The pellet does not now affect
the sound of the whistle made by blowing across a small vent.
SUMMARY
The preceding comments relate primarily to the utilization of
the Lagenaria siceraria identified by Dr. Whitaker. Whether,
in addition, they were ever eaten might be determined by an
examination of the coprolites collected. The Cucurbita, on the
other hand, were obviously an important food. The significance
of their presence in this early horizon remains to be evaluated,
and the explanation will more likely be supplied by botanists than
by archaeologists. Actually we have no clue as to the original
home of these people, and unfortunately the nature of their
artifacts is such as to promise great difficulty in tracing them.
Hence such contributions as Dr. Whitaker's botanical analysis are
most welcome.
AMERICAN MUSEUM NOVITATES
8
NO. 1426
BOTANICAL IDENTIFICATION AND SIGNIFICANCE
Through the courtesy of Mr. Junius Bird, the writer has been
privileged to examine and study the remarkably well-preserved
cucurbitaceous materials excavated at Huaca Prieta. The wellorganized, systematic excavations directed by Mr. Bird have
uncovered some very significant plant material, and a preliminary
report (Bird, 1948) has established the fact that the peoples
occupying this site were among the earliest agriculturists in the
New World. The cucurbit materials are of such importance
for ethnobotanists, archaeologists, and others that a separate
report of their analysis and identification seems justified.
METHODS
For the most part the cucurbit material submitted for examination was in an excellent state of preservation, considering its age.
The shells' of both Lagenaria and Cucurbita were in good condition, and in some instances the fruits of Lagenaria were recovered
virtually intact (fig. 1). The seeds were not so well preserved
as the shells. For some specimens, erosion of the outer seed
coat and general deterioriation through age made positive identification impossible.
A method for softening and decolorizing the tissue of the walls
of the shells and mounting it for microscopic examination has
been described previously (Whitaker, 1948). Whenever a particular fragment could not be positively identified as either Lagenaria or Cucurbita by macroscopic means, it was sectioned and
examined microscopically. The histological pattern of the walls
of the shell is so different for the two genera that rapid and positive
identification is a routine matter.
The seeds of Lagenaria siceraria (Molina) Standley are quite
distinctive and can easily be identified by inspection. Seeds of
Cucurbita were checked with the extensive collection maintained
1
In a previous paper, the author (Whitaker, 1948) used the term "exocarp" to
designate the remains of the fruit, i.e., the so-called "shells" or "rind." Dr. R. M.
Brooks, Associate Pomologist, University of California, informs me that this term is
not applicable to fruit with an inferior ovary, where the carpels are surrounded by
accessory tissue. According to Dr. Brooks, the morphological nature of the mature
pepo has not been satisfactorily worked out, and for this reason a meaningful term
to describe the hard, dry, tough, and brittle remains of the fruit has not been suggested. In the meantime the word "shell" has been applied to these fragments, with
the understanding that it has no status in botanical terminology.
1949
CUCURBIT MATERIALS FROM PERU
9
at this laboratory. By comparison it was possible to determine
rather accurately the precise affiliation of the archaeological seeds.
MATERIALS
A list of the cucurbit materials follows, with notes on their
identification and other pertinent comments:
PRE-CERAMIC, PRE-MAIZE GOURDS AND SQUASH; HUACA PRIETA, TEST 3
HP-3-M
81 stems
This collection of peduncles appears to be chiefly those of some species
of Cucurbita, probably C. ficifolia, Bouche. There are a few that may be
Lagenaria siceraria, but certainly most of them are Cucurbita
62 roughly spherical shells with stem scars
Evidently most of this material is some species of Cucurbita, probably
C. ficifolia. Shells range in thickness from 2 mm. to 3 mm. A majority
of the fragments have peduncle scars. However, about one-third are
from the distal portion of the fruit with the scar and depression from the
withered flower parts
11 thin shells, like the preceding
All except 2 of these shells are almost certainly those of some species of
Cucurbita, probably C. ficifolia. The two exceptions have a wall thickness of 3.5 mm. and 4 mm., respectively, and when sectioned proved to be
Lagenaria siceraria. The remainder have walls averaging about 2 mm.
in thickness
12 bottle gourd necks
This collection consists of 12 fragments of the neck portion of the fruits
of Lagenaria siceraria. The fruits were evidently of the short-necked
type; 6 of the fragments retain vestiges of the peduncle. In one case,
the portion of the peduncle remaining attached to the neck fragment
was approximately 10 mm. in length. The shells are of varying thickness (4 mm. to 8 mm.)
5 centers of base
Fragments from distal portion of the fruit, with depressed area and scar
from withered flower parts. Two are an unidentified species of Cucurbita
(these were sectioned), wall thickness average, 2.5 mm.; the remainder
are Lagenaria siceraria, wall thickness, 5 to 6 mm.
33 samples of gourd shell showing range in thickness
This collection consists of 33 fragments of shells of Lagenaria siceraria.
The range in shell thickness for the entire group varies from 4 mm. to 9
mm.; in area from approximately 150 sq. mm. to 2700 sq. mm.
Seeds as listed on field analysis:
12 seeds of Lagenaria siceraria
The seeds in this collection show some of the extreme broadness and
size characteristic of those in collections of these species from the Old
World (fig. 3D)
AMERICAN MUSEUM NOVITATES
10
NO. 1426
SI
FIG. 3. Seeds of cucurbits from Huaca Prieta, showing range of size and
shape of mature seeds. A. Cucurbita ficifolica. B. C. moschata, fringed-margined variety. C. Lagenaria siceraria, seeds taken from net float (fig. iB).
Note small size and absence of paired protuberances. D. L. siceraria, broad,
flat type of seeds characteristic of the Old World collections of this species.
Millimeter scale at bottom.
18 seeds
These are the seeds of the fig-leaf gourd, Cucurbita ficifolia (fig. 3A)
1 large batch of seeds
These are mostly seeds of an unidentified species of Cucurbita, probably
C. moschata, Duchesne, since some of them have a fringed margin (fig.
3B). There is a single seed that appears to be of C. ficifolia. There are
also a few of Lagenaria siceraria
HP-3-G
1 bottle gourd net float and seeds removed from it
Lagenaria siceraria, relatively long-necked type; over-all length, 242
mm.; length of neck, 135 mm.; maximum diameter of bulbous portion,
120 mm. Peduncle end corked with tightly fitted plug apparently of
same material. Seeds typical of L. siceraria found in archaeological
collections from South America: average length, 13 mm.; average
width, 6 mm. (see fig. 3C)
1^949
CUCURBIT MATERIALS FROM PERU
11
GOURDS AND SQUASH ASSOCIATED WITH CORNCOBS AND CUPISNIQUE POTTERY,
HUACA PRIETA, TEST 5
HP-5-A1
1 fragmented gourd bowl
Shell 6 mm. thick
6 gourd fragments
Shells ranging from 5 mm. to 10 mm. in thickness
2 stems
1 peduncle 20 mm. in length; the other with a portion of the shell adhering to the peduncle, about 12 mm. long. All the above items are undoubtedly Lagenaria siceraria
HP-5-B
2 bottle gourd necks
Neck portion of Lagenaria siceraria fruits; one 30 mm. long with shell
3 mm. thick; the other 37 mm. long with shell 5 mm. thick
1 bottle gourd container
Lagenaria siceraria, fragmented, evidently a rather small or immature
fruit, the remaining portion appears to be about one-half of the original
container; shell wall, 5 mm. thick
1 bottle gourd bowl rim fragment
Lagenaria siceraria, apparently from rim of container used as a bowl;
area, 2200 sq. mm., shell wall, 5 mm. thick
6 warty squash fragments
Shell fragments of some warty Cucurbita, almost certainly C. moschata
7 stems
Peduncles of Lagenaria siceraria and unidentified species of Cucurbita
HP-5-B8
19 stems
Peduncle fragments; all but 1 are classified as Cucurbita ficifolia The
single exception is a peduncle with a portion of the shell attached. Since
the attached shell is warty, it is probably some species of Cucurbita, most
likely C. moschata
7 warty squash fragments and 2 shells with stem scars
9 shell fragments of Cucurbita sp.; 6 are warty, 3 are smooth; one of the
latter may be Lagenaria siceraria. The warty shells are probably those of
C. moschata and are from 1.5 mm. to 3.5 mm. in thickness. The smooth
shells vary from 1.5 mm. to 3 mm. in wall thickness
HP-5-B8
Large collection of seeds of Cucurbita sp.
Most of them badly eroded, and damaged, perhaps by animals. One
seed is definitely C. moschata; the remainder are probably C. ficifolia
HP-5-House 5, Huaquero pit at entrance
1 bottle gourd
Much the same type as that used for net float (see fig. 1A); over-all
length, 245 mm.; neck length, 145 mm.; maximum diameter of bulbous
portion, 115 mm.; wall, 4 mm. thick
1 gourd bowl
Broken, fragmented, large; maximum diameter, 173 mm.; wall, 6 mm.
thick: Lagenaria siceraria
12
AMERICAN MUSEUM NOVITATES
NO. 1426
SQUASH SEED AND GOURD SEED ASSOCIATED WITH POTTERY OF CUPISNIQUE AND
PRE-CUPISNIQUE TYPE
HP-5, between Floors 1 and 2 of House 7
Cucurbita ficifolia
HP-5-A, general
2 seeds of Lagenaria siceraria; 1 small immature seed of Cucurbita
moschata
HP-5, House 2
1 of Lagenaria siceraria; 1 Cucurbitaficifolia; 1 C. moschata
HP-5, House 5
In poor condition, many appear to be immature; 7 can be positively
identified as Cucurbita moschata; the remainder are C. ficifolia
HP-5-A2
7 seeds of Cucurbitaficifolia; 3 of C. moschata; 3 of Lagenaria siceraria
HP-5-A3
2 seeds of Lagenaria siceraria; 5 of Cucurbita moschata; the remainder are
C. ficifolia
HP-5-A4
1 seed of Cucurbita ficifolia and several immature seeds of unidentifiable
species of Cucurbita, possibly C. moschata; the remainder are Lagenaria
siceraria
HP-5-B1
2 seeds of Cucurbita ficifolia
HP-5-B6
1 seed of Lagenaria siceraria, and 1 seed of unidentified Cucurbita badly
deteriorated but probably C.ficifolia
HP-5-B1O
Mostly in poor condition; 6 seeds of Cucurbita moschata; the remainder
C. ficifolia
HP-5-B11
1 seed of Lagenaria siceraria and 1 of Cucurbita ficifolia
HP-5-C4
2 seeds of Lagenaria siceraria; the remainder are in poor condition but
are evidently some species of Cucurbita; at least 4 are C. ficifolia and 3 are
C. moschata
HP-5-D1
1 seed of Cucurbitaficifolia
HP-5-D2
1 seed of Lagenaria siceraria; 3 seeds of Cucurbita ficifolia; remainder
small seeds with fringed margin, C. moschata
DISCUSSION AND CONCLUSIONS
The material from the pre-ceramic, pre-maize horizon at
Huaca Prieta indicates that Lagenaria siceraria and two species of
Cucurbita (C. ficifolia and C. moschata) were being cultivated during this period. The abundant materials of L. siceraria suggest
that it was widely used for a number of purposes by the inhabi-
1949
CUCURBIT MATERIALS FROM PERU
13
tants of this site (see p. 3). The large, broad seeds (see fig.
3D) show a striking resemblance to the Old World collections of
this species. This point may or may not have some significance.
However, it does seem important to note that they are markedly
unlike the small narrow seeds, mostly without paired protuberances, characteristic of other South American archaeological
collections of this species (compare figs. 3C and 3D). This new
information suggests that if L. siceraria is a native of the Old
World, it was introduced into South America at a very early
date. On the other hand, there is no clear evidence that negates
the idea that this species was a New World endemic and was
domesticated in the Americas.
There is good evidence from the peduncles, shell fragments, and
seeds that Cucurbita ficifolia was cultivated contemporaneously
with Lagenaria. In addition, there is some indication from the
seeds (one large batch of seeds) that another species of Cucurbita
was present. The fringed margins of these specimens permits
one with some confidence to assign them to C. moschata (Whitaker
and Bohn, in press). In this group are found the modern varieties
Sugar Marvel, Butternut, etc. However, at this horizon no substantiating evidence from peduncles or shells has been discovered.
Thus from the pre-ceramic, pre-maize horizon there is evidence
for the presence of three species of the family Cucurbitaceae;
first, abundant material of shells, peduncles, and peculiar broad
seeds of Lagenaria siceraria; second, peduncles, seeds, and possibly shells of Cucurbita ficifolia; third, seeds of C. moschata.
The estimated date for the beginning of this horizon is about
3000 B.C. (Bird, 1948).
Coming to the corncob, Cupisnique pottery horizon, Lagenaria
siceraria is still present in abundance, but there are no specimens
of the broad-seeded form found in the earlier horizon. Cucurbita
ficifolia continued to be cultivated, although the seeds of this
species are somewhat different from those associated with the
pre-ceramic, pre-maize horizon in that they seem to be specimens
from the dingy white-seeded form of this species rather than the
black-seeded form characteristic of the older horizon.
A new element, the warty squash, enters the picture at this
point, occurring stratigraphically slightly earlier than maize. It
is probably a form of Cucurbita moschata. Warty varieties are
found in C. moschata, C. pepo, and C. maxima, but warty forms of
14
AMERICAN MUSEUM NOVITATES
NO.
1426
C. ficifolia are unknown. Cucurbita pepo is strictly North
American in origin and distribution. Thus C. moschata and
C. maxima remain as possibilities. There is considerable evidence
that C. moschata is endemic in the northern portion of South
America, i.e., Peru and Colombia, while C. maxima is native
to southern Brazil and northern Argentina. Additional evidence
is supplied by the seeds which are typical of those with a fringed
margin characterizing some varieties of C. moschata (Whitaker
and Bohn, in press). Furthermore, none of the seeds in the entire collection resembled C. maxima.
It is probable that the variety of C. moschata cultivated at this
time had a warty exterior and seeds with a pronounced, fringed
margin. However, it must be remembered that warty shells were
not associated with fringed-margined seeds in the earlier horizon.
Varieties of Cucurbita moschata with fringed-margined seeds
must have had a widespread distribution in Peru during pre-contact times. Carter (1945) reports seeds of this type in the Uhle
collections from Chincha, Peru. The chronology is given as
late pre-Columbian (ca. 1300-1500 A.D.).
If Cucurbita ficifolia was cultivated by the peoples of this
period, and the evidence is very good in this respect, its presence
deserves special comment. Cucurbita ficifolia has never been
reported from archaeological collections in either North or South
America. It has generally been considered to be endemic in
Mexico and Central America. Its presence in Peru at this
early date is a puzzle, unless we assume that it is widely adapted.
If so, it could have been carried northward by a succession
of early migrations.
To summarize, the most important facts brought out by
examination of the cucurbitaceous materials from Huaca Prieta
are: 1, there is abundant material of Lagenaria siceraria present
in both horizons; 2, in the pre-ceramic, pre-maize horizon there
are seeds of the broad-seeded form of L. siceraria, typical of collections of this species from the Old World; 3, a black-seeded
form of Cucurbita ficifolia is present in the early horizon, and a
dingy white-seeded form in the later horizon; 4, in the preceramic, pre-maize horizon, there are fringed-margined seeds of
C. moschata; 5, in the later horizon (corncob, Cupisnique pottery) there are warty shells of some variety of C. moschata; associated with these warty shells are fringed-margined seeds of the
same species.
1949
CUCURBIT MATERIALS FROM PERU
15
LITERATURE CITED
BENNETT, WENDELL C.
1948. A reappraisal of Peruvian archaeology. Mem. Soc. Amer. Archaeol.,
no. 4, Amer. Antiquity, vol. 13, no. 4, pt. 2.
BIRD, JUNIus B.
1948. America's oldest farmners. Nat. Hist., vol. 57, pp. 296-303, 334-335.
CARTER, G. F.
1945. Some archeologic cucurbit seed from Peru. Acta Amer., vol. 3, pp.
163-172.
WHITAKER, T. W.
1948. Lagenaria: a pre-Columbian cultivated plant in the Americas.
Southwestern Jour. Anthrop., vol. 4, pp. 49-58.
WHITAKER, T. W., AND G. W. BOHN
[In press.] The taxonomy, genetics, production and uses of the cultivated
species of Cucurbita.